DESCRIPTION: Localization of epileptogenic zones in epilepsy patients is a prerequisite for the surgical relief of medically refractory partial seizures. In patients with non-lesional extratemporal epilepsy (ETE), localization of seizure foci is usually difficult with traditional non-invasive methods such as scalp electroencephalography, positron emission tomography, single photon emission computed tomography, magnetic resonance imaging and neurological testing. The long-term goal of the proposed study is to establish proton metabolite mapping with chemical shift imaging as a means of localization for seizure foci in ETE patients. To this end, a number of difficulties will need to be tackled including signal contamination from subcutaneous lipid, partial volume effects and difficulty in accessing patient data without absolute quantification of the metabolites. The specific aims of this project have not changed except for the specific inclusion of the patient studies. They are: 1) to develop a new technique for obtaining proton chemical shift images of the head without signal contamination from the subcutaneous lipid, 2) to obtain partial volume corrected measurements of relative metabolite levels in normal volunteers in order to establish a reference data base of metabolites of relevance, 3) to develop and validate an adaptive normalization technique that takes into account variations and relative metabolite levels in normal controls and in individual subject anatomy and, 4) to validate the sensitivity of the proposed development in a population of patients with known seizure foci. This work will be carried out on a 1.5 Tesla whole body magnetic resonance imaging/spectroscopy system. For specific aim 1, improvements will be implemented to the 2DCSI spin warp technique. This will involve the development of echoplaner spectroscopic imaging combined with a conventional CSI sequence to cover k-space in a hybrid fashion. For specific aim 2, correction of the partial volume effect will be approached utilizing segmentation of anatomic images. The first component of the correction will be incorporation of partial volume effect in the derivation of the tissue specific relative metabolite levels in the brain. Secondly, correction for individual anatomy in the normalization of the patient data will be attempted. Over the three year period of the proposal the sequence development and validation will be completed in the first year. Development and implementation of processing algorithms will start concurrently with the sequence development and will be validated with experimental data in the early part of the second year. Patient studies will be conducted in the second and third years of the proposal to validate the technical developments and establish clinical protocols suitable for more extensive studies in the future.